Anthropic Warns It May Near AI-Driven Progress Threshold

Anthropic’s filing says internal estimates may soon meet its policy threshold for AI-driven progress doubling, while stating no current model has crossed it. Fable 5.2 and Mythos 5.2 launched amid slowdown and IPO discussions, but performance, safety, and progress claims are not independent evidence of self-building AI. Watch for policy changes and corroboration.